- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
[Lecture Notes in Computer Science] Neural Information Processing Volume 11306 (25th International Conference, ICONIP 2018, Siem Reap, Cambodia, December 13–16, 2018, Proceedings, Part VI) || Fast Image Recognition with Gabor Filter and Pseudoinverse Learning AutoEncoders
摘要： Deep neural network has been successfully used in various ?elds, and it has received signi?cant results in some typical tasks, especially in computer vision. However, deep neural network are usually trained by using gradient descent based algorithm, which results in gradient vanishing and gradient explosion problems. And it requires expert level professional knowledge to design the structure of the deep neural network and ?nd the optimal hyper parameters for a given task. Consequently, training a deep neural network becomes a very time consuming problem. To overcome the shortcomings mentioned above, we present a model which combining Gabor ?lter and pseudoinverse learning autoencoders. The method referred in model optimization is a non-gradient descent algorithm. Besides, we presented the empirical formula to set the number of hidden neurons and the number of hidden layers in the entire training process. The experimental results show that our model is better than existing benchmark methods in speed, at same time it has the comparative recognition accuracy also.
关键词： Pseudoinverse learning autoencoder,Gabor ?lter,Handcraft feature,Image recognition
更新于2025-09-23 15:19:57
-
Development of a Recognition System for Spraying Areas from Unmanned Aerial Vehicles Using a Machine Learning Approach
摘要： Unmanned aerial vehicle (UAV)-based spraying systems have recently become important for the precision application of pesticides, using machine learning approaches. Therefore, the objective of this research was to develop a machine learning system that has the advantages of high computational speed and good accuracy for recognizing spray and non-spray areas for UAV-based sprayers. A machine learning system was developed by using the mutual subspace method (MSM) for images collected from a UAV. Two target lands: agricultural croplands and orchard areas, were considered in building two classifiers for distinguishing spray and non-spray areas. The field experiments were conducted in target areas to train and test the system by using a commercial UAV (DJI Phantom 3 Pro) with an onboard 4K camera. The images were collected from low (5 m) and high (15 m) altitudes for croplands and orchards, respectively. The recognition system was divided into offline and online systems. In the offline recognition system, 74.4% accuracy was obtained for the classifiers in recognizing spray and non-spray areas for croplands. In the case of orchards, the average classifier recognition accuracy of spray and non-spray areas was 77%. On the other hand, the online recognition system performance had an average accuracy of 65.1% for croplands, and 75.1% for orchards. The computational time for the online recognition system was minimal, with an average of 0.0031 s for classifier recognition. The developed machine learning system had an average recognition accuracy of 70%, which can be implemented in an autonomous UAV spray system for recognizing spray and non-spray areas for real-time applications.
关键词： image classifiers,machine learning system,precision agriculture,recognition system,mutual subspace method
更新于2025-09-19 17:15:36
-
Iris Recognition Using Gauss Laplace Filter
摘要： Biometrics deals with recognition of individuals based on their behavioral or biological features. The recognition of IRIS is one of the newer techniques of biometrics used for personal identification. It is one of the most widely used and reliable technique of biometrics. In this study a novel approach is presented for IRIS recognition. The proposed approach uses Gauss Laplace filter to recognize IRIS. The proposed approach decreases noise to the maximum extent possible, retrieves essential characteristics from image and matches those characteristics with data in a database. This method will be effective and simple and can be implemented in real time. The experiments are carried out using the images of IRIS acquired from a database and MATLAB application has been applied for its effective and simple manipulation of IRIS image. It was observed that developed approach has more accuracy and a relatively quicker time of execution than that of the existing approaches.
关键词： IRIS Recognition,Biometrics,Gauss Laplace Filter
更新于2025-09-19 17:15:36
-
DNA-MnO2 Nanosheets as Washing- and Label-Free Platform for Array-Based Differentiation of Cell Types
摘要： Accurate and facile differentiation of cell types is critical for accurate diagnosis and therapy of diseases. However, it remains challenging due to low specificity, requirement of sophisticated instruments, and tedious operation steps. Herein, a simple, washing- and label-free chemical tongue was constructed for differentiation of cell types. In the array-based sensing platform, DNA-ligand ensembles adsorbed on the surface of MnO2 nanosheets were used as sensing probes. Instead of aptamers from cell-SELEX, the randomly designed DNA strands were used, offering versatile interactions with cells. The property that MnO2 nanosheets can be degraded by intracellular glutathione makes the platform avoid the washing step. Eight types of cell lines were distinguished from each other after the data were treated with principal component analysis (PCA). In addition, a 95% of identification accuracy for the randomly selected unknown samples was achieved. The strategy shows an excellent performance not only in distinguishing cell lines but also in the identification of unknown cell samples.
关键词： label-free,pattern recognition,cell types,DNA-MnO2 nanosheets,washing-free
更新于2025-09-19 17:15:36
-
A novel surface plasmon resonance sensor based on a functionalized graphene oxide/molecular-imprinted polymer composite for chiral recognition of <scp>l</scp> -tryptophan
摘要： Herein, a novel surface plasmon resonance (SPR) sensor based on a functionalized graphene oxide (GO)/molecular-imprinted polymer composite was developed for the chiral recognition of L-tryptophan (L-Trp). The composite's recognition element was prepared via a facile and green synthesis approach using polydopamine as both a reducer of GO and a functional monomer as well as a cross-linker for molecular imprinting. The composite was characterized via Fourier transform infrared spectroscopy, scanning electron microscopy, X-ray diffraction, and Raman spectroscopy. After attaching the composite onto the gold surface of an SPR chip, the sensor was characterized using contact-angle measurements. The sensor exhibited excellent selectivity and chiral recognition for the template (i.e., L-Trp). Density functional theory computations showed that the difference in hydrogen bonding between the composite element and L-Trp and D-Trp played an important role in chiral recognition.
关键词： molecular-imprinted polymer,graphene oxide,surface plasmon resonance,L-tryptophan,chiral recognition
更新于2025-09-19 17:15:36
-
Reconstructing 3D Face Models by Incremental Aggregation and Refinement of Depth Frames
摘要： Face recognition from two-dimensional (2D) still images and videos is quite successful even with “in the wild” conditions. Instead, less consolidated results are available for the cases in which face data come from non-conventional cameras, such as infrared or depth. In this article, we investigate this latter scenario assuming that a low-resolution depth camera is used to perform face recognition in an uncooperative context. To this end, we propose, first, to automatically select a set of frames from the depth sequence of the camera because they provide a good view of the face in terms of pose and distance. Then, we design a progressive refinement approach to reconstruct a higher-resolution model from the selected low-resolution frames. This process accounts for the anisotropic error of the existing points in the current 3D model and the points in a newly acquired frame so that the refinement step can progressively adjust the point positions in the model using a Kalman-like estimation. The quality of the reconstructed model is evaluated by considering the error between the reconstructed models and their corresponding high-resolution scans used as ground truth. In addition, we performed face recognition using the reconstructed models as probes against a gallery of reconstructed models and a gallery with high-resolution scans. The obtained results confirm the possibility to effectively use the reconstructed models for the face recognition task.
关键词： anisotropic error,3D reconstruction,3D face recognition,Depth data
更新于2025-09-19 17:15:36
-
Improved triangular-based star pattern recognition algorithm for low-cost star trackers
摘要： Star identification algorithms based on triangular-pattern are more suitable for low-cost star trackers since they require less star density in the field of view to operate effectively. In this paper, we propose a modified star pattern recognition algorithm based on the triangular-based algorithm of “LIEBE”. The main contribution of the proposed work is twofold. First, a new strategy for the selection of star triplets is proposed for database construction. Second, new selection criteria of the reference star are considered for pattern generation process. A sky simulation program is developed to assess mainly the robustness against different conditions of noise. The obtained results show an improvement in the overall identification rate, more robustness towards missing stars, and more efficiency towards magnitude noise. Furthermore, our proposed algorithm shows comparable robustness with the recently proposed triangular algorithms despite their reliance on more accurate camera and a validation process. To assess the algorithm performance, the algorithm is implemented on a prototype of Data Processing Unit (DPU) based on ARM Cortex-M4 processor. In this part, we discuss the major design decisions and we present the hardware architecture of DPU. The algorithm shows promising running time at a reduced on-board database when implemented on ARM platform.
关键词： Small satellite,Star tracker,Hardware implementation,Star pattern recognition,Star database optimization,Star identification
更新于2025-09-19 17:15:36
-
Document Verification: A Cloud-Based Computing Pattern Recognition Approach to Chipless RFID
摘要： In this paper, we propose a novel means of verifying document originality using chipless RFID systems. The document sender prints a chipless RFID tag into the paper and does a frequency scanning in the 57–64 GHz spectrum of the document. The results of scattering parameters in individual step frequencies are stored in a cloud database, denoised and passed to pattern classi?ers, such as support vector machines or ensemble networks. These supervised learners train themselves based on these data on the remote/cloud computer. The document receiver veri?es this frequency ?ngerprint by using the same scanning method, sending the scattering parameters to the cloud server and getting the decoded data. Paper originality is veri?ed if the decoded data are as expected. The advantages of our cloud chipless RFID processing deployments are cost reduction and increased security and scalability.
关键词： chipless tag,classi?cation algorithms,Radio frequency identi?cation,support vector machines,pattern recognition,cloud computing,ensemble networks
更新于2025-09-19 17:15:36
-
[IEEE 2018 Eighth International Conference on Image Processing Theory, Tools and Applications (IPTA) - Xi'an, China (2018.11.7-2018.11.10)] 2018 Eighth International Conference on Image Processing Theory, Tools and Applications (IPTA) - Deep Dilated Convolutional Network for Material Recognition
摘要： Material is actually one of the intrinsic features for objects, consequently material recognition plays an important role in image understanding. For the same material, it may have various shapes and appearances, but keeps the same physical characteristic, which brings great challenges for material recognition. Most recent material recognition methods are based on image patches, and cannot give accurate segmentation results for each specific material. In this paper, we propose a deep learning based method to do pixel level material segmentation for whole images directly. In classical convolutional network, the spacial size of features becomes smaller and smaller with the increasing of convolutional layers, which loses the details for pixel-wise segmentation. Therefore we propose to use dilated convolutional layers to keep the details of features. In addition, the dilated convolutional features are combined with traditional convolutional features to remove the artifacts that are brough by dilated convolution. In the experiments, the proposed dilated network showed its effectiveness on the popular MINC dataset and its extended version.
关键词： Dilated convolution network,Material recognition
更新于2025-09-19 17:15:36
-
[IEEE 2018 Eighth International Conference on Image Processing Theory, Tools and Applications (IPTA) - Xi'an, China (2018.11.7-2018.11.10)] 2018 Eighth International Conference on Image Processing Theory, Tools and Applications (IPTA) - A multiple classifiers-based approach to palmvein identification
摘要： The usual trend for the conventional palmvein recognition techniques is first to extract discriminative hand-crafted feature representations from the raw images, and then feed a classifier with them. Unfortunately, it is not yet clear how the effectiveness of such features may be held in case of a large user population or in environments where the variability among acquisitions of the same person may increase. In order to face with this problem, it may be considered that the use of multiple classifiers may increase the recognition performance with respect to that of the best individual classifier, and also may handle the problem of an effective feature extraction step. In this paper, we explore the ensemble classifier approach based on Random Subspace Method (RSM), where the basic feature space is derived after a preliminary feature reduction step on the source image, and compare results achieved with and without the use of hand-crafted features. Experimental results allow us concluding that this approach leads to better results under different environmental conditions.
关键词： Image processing,palmvein recognition,features,multiple classifiers
更新于2025-09-19 17:15:36